IRC log started Sat Nov 20 00:00:00 1999
[msg(TUNES)] permlog 1999.1120
-:- SignOff air: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
-:- Fufie [stig@tunnel-44-10.vpn.uib.no] has joined #tunes
-:- SignOff Fufie: #TUNES (Leaving)
-:- Connection closed from irc.us.openprojects.net: Success
-:- Connecting to port 6667 of server irc.us.openprojects.net [refnum 0]
-:- BitchX+Deb1an: For more information about BitchX type /about
-:- Welcome to the Internet Relay Network TUNES (from carter.openprojects.net)
-:- Your host is carter.openprojects.net, running version u2.10.05.18.(ipcheck4-5) (from carter.openprojects.net)
-:- This server was cobbled together Wed Apr 28 1999 at 12 02:19 EDT(from carter.openprojects.net)
-:- carter.openprojects.net u2.10.05.18.(ipcheck4-5) dioswkfcg biklmnopstv 
-:- [local users on irc(37)] 7%
-:- [global users on irc(204)] 41%
-:- [invisible users on irc(295)] 59%
-:- [ircops on irc(14)] 3%
-:- [total users on irc(499)]
-:- [unknown connections(0)]
-:- [total servers on irc(33)] (avg. 15 users per server)
-:- [total channels created(153)] (avg. 3 users per channel)
!carter.openprojects.net Highest connection count: 70 (69 clients)
!carter.openprojects.net Welcome to Open Projects!  You are on 3 ca 1(2) ft 14(14) tr.
-:- Mode change [+f] for user TUNES
-:- Mode change [+iws] for user TUNES
-:- JOIN activated by "TUNES #tunes tunes@bespin.dhs.org "
-:- TUNES [tunes@bespin.dhs.org] has joined #tunes
-:- Topic for #TUNES: TUNES -- A free reflective computing environment. Thought and Discussion. Implementation too some day..
-:- topic set by Fufie [Thu Nov 18 11:19:20 1999]
-:- [Users(#tunes:2)] 
[ TUNES     ] [ abi       ] 
-:- Channel #tunes was created at Sun Feb 28 08:48:06 1999
-:- BitchX+Deb1an: Join to #tunes was synced in 6.902 secs!!
-:- Mode change [-s] for user TUNES
-:- Fare [fare@quatramaran.ens.fr] has joined #Tunes
<Fare> forcer: interested in helping?
06:20am
-:- athanor [toto@MS151U23.U-3MRS.FR] has joined #Tunes
<athanor> hello y a quelqu'un ?
07:00am
-:- SignOff athanor: #TUNES (changing servers)
-:- athanor [toto@MS151U23.U-3MRS.FR] has joined #tunes
-:- SignOff athanor: #TUNES (Leaving)
-:- SignOff Fare: #TUNES (Ping timeout for Fare[quatramaran.ens.fr])
-:- Fare [fare@quatramaran.ens.fr] has joined #Tunes
-:- _ruiner_ [blah@ppp085.wi.centurytel.net] has joined #tunes
-:- SignOff _ruiner_: #TUNES (Leaving)
-:- fire [no@209-68-229-41.dialup.cust.tfb.com] has joined #tunes
-:- SignOff fire: #TUNES (Ping timeout for fire[209-68-229-41.dialup.cust.tfb.com])
-:- fire`` [no@209-68-229-77.dialup.cust.tfb.com] has joined #tunes
-:- fire`` is now known as fire
<fire> i guess it's just us and the bots now
10:30am
<fire> what happened to all the people
10:40am
-:- eihrul [lee@usr5-ppp130.lvdi.net] has joined #tunes
<fire> hi
<fire> anyone noticed that the last update was four months ago?
10:50am
-:- yoo [ultima@user-38lc64q.dialup.mindspring.com] has joined #Tunes
* yoo/#Tunes bashes eihrul
<fire> howdy y00
<fire> *sigh*
<fire> there is still a small flame known as TUNES
<fire> not yet YAOS
<yoo> heh
<eihrul> yo yoo
11:00am
* yoo/#Tunes pokes eihrul
-:- SignOff yoo: #TUNES (Leaving)
-:- SignOff fire: #TUNES (Leaving)
-:- yoo [ultima@user-38lc6aj.dialup.mindspring.com] has joined #Tunes
<eihrul> re
<yoo> eihrul
-:- hcf [nef@me-portland-us938.javanet.com] has joined #tunes
11:40am
<Fare> hi
<yoo> Fare!
<Fare> ult
<yoo> How ya doin man! *shakes Fare's hand*
* Fare/#Tunes deep-sixes yoo
-:- FareWell [fare@quatramaran.ens.fr] has joined #Tunes
-:- FareWell [fare@quatramaran.ens.fr] has left #Tunes []
11:50am
-:- SignOff hcf: #TUNES (Leaving)
-:- ProGuy [paul@p449-51.ppp.get2net.dk] has joined #tunes
-:- SignOff yoo: #TUNES (Leaving)
-:- pyro [tcn@cci-209150250039.clarityconnect.net] has joined #tunes
<pyro> knock knock
<ProGuy> hey
-:- AlonzoTG [Alonzo@216-164-130-51.s559.tnt1.lnhva.md.dialup.rcn.com] has joined #tunes
03:20pm
<pyro> oh hey
<pyro> hey these words are wrong!
<ProGuy> ?
03:30pm
<pyro> jk
<ProGuy> oh
<pyro> punk songs are different every time :)
<ProGuy> ha
<pyro> anyway.. time to get my OS running again
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
03:40pm
-:- SignOff pyro: #TUNES (pyro has no reason)
-:- rares [rares@wtrb-sh4-port189.snet.net] has joined #tunes
-:- nachas [user7166@user-2iveabo.dialup.mindspring.com] has joined #tunes
-:- nachas [user7166@user-2iveabo.dialup.mindspring.com] has left #tunes []
-:- mibin [mibin@62.10.7.155] has joined #tunes
-:- SignOff ProGuy: #TUNES (leaving..., remember http://tpp.dhs.org)
-:- hcf [nef@me-portland-us734.javanet.com] has joined #tunes
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp130.lvdi.net])
-:- eihrul [lee@usr5-ppp59.lvdi.net] has joined #tunes
<mibin> abi, glycerine
<mibin> abi, glycerine?
<abi> i haven't a clue, mibin
<mibin> abi, methane?
<abi> i haven't a clue, mibin
05:50pm
-:- Kaufmann [Kaufmann@200.224.105.180] has joined #tunes
<Kaufmann> Shalom!
<Kaufmann> Ma sh'lom ha?
<Kaufmann> Anyone?
06:00pm
<eihrul> hey
<Kaufmann> Hey
<Kaufmann> Free to talk?
<eihrul> sure
<Kaufmann> I'm working on a completely modular, reflective Perl/Tk program for networking functions
<Kaufmann> Like browsers, email programs, etc.
<mibin> reflective browsers?
<eihrul> how will this be helpful at all? :)
<mibin> how will tunes be helpful at all?
<Kaufmann> mibin, yep
<Kaufmann> LOL mibin
<eihrul> chances are, the networking package is going to be reflective
<eihrul> but the rest is not
<eihrul> and most people don't mess with networking
<eihrul> so the sum of its parts in this case is not greater than the other whole :)
<Kaufmann> eihrul, for one, the program will have a floating window into which you can type Perl and access internal data structures (such as the properties of each module, etc.)
<Kaufmann> "... you can type Perl CODE", even
<eihrul> we have emacs, what do we need that for?
<Kaufmann> well, it's pretty :)
<Kaufmann> Seriously, the focus is different from emacs'
<eihrul> i don't see how...
<eihrul> plus emacs does everything at this point
<Kaufmann> Oh well
<Kaufmann> Which is a good thing?
<eihrul> no :)
<eihrul> i just said it does everything...
<eihrul> Kaufmann: there's nothing wrong with making perlmacs or anything :)
<Kaufmann> Well, it's just nice to have a program into which you can quickly hack up a graphical module to do whatever task you want to do, save it, run it (by do()) and use it, all in one shot
06:10pm
<Kaufmann> And of course, the core will be light
<Kaufmann> Plus, it'll be relatively easy to add a function to load modules from remote servers (through HTTP or whatever), just as if you were loading them from local files
<Kaufmann> ...which will be /nice/.
<eihrul> ...tunes will be /nice/.
<mibin> both vaporware
<Kaufmann> eihrul, of course TUNES will be nice. But mine will be done before the age of Capricorn. (Hopefully.) :)
<Kaufmann> mibin, /I/'ve already started coding
<mibin> ah, its ok then
<mibin> i am sorry i am not in the best mood 2day
<eihrul> Kaufmann: there was preliminary tunes code before too...
<eihrul> so doesn't prove much :)
<rares> libglade
<Kaufmann> The core is pretty much a very simple program... load top-level widgets and subs, load modules from /modules dir into Module objects in %modules, main loop. All the actual functionality is in the module code.
<rares> uses gtkML
<rares> K that's what you need
<Kaufmann> rares, why on Earth? I'm using Perl/Tk.
<rares> what you is a quick way to create a user interface right
<rares> you want even
<Kaufmann> not really
06:20pm
<rares> okay then what do you want :)
<Kaufmann> What I want is a system where the user can do whatever he wants to do in a natural way, by customizing the program to suit his own needs
<eihrul> Kaufmann: perl is natural?
<rares> without recompiling?
<eihrul> Kaufmann: you're not targetting at users, just perl programmers
<rares> @_!# @ is not natural
<rares> we need UMLshell not GUIs
<Kaufmann>  eihrul, to me it is :) Seriously, you will only have to use Perl if you want to do complex stuff like altering modules, changing the UI's behaviour, etc... you can always just point and click, that's why it's graphical in the first place.
<Kaufmann> rares, yes, without recompiling
<rares> that's an implementation problem
<Kaufmann> rares, actually, when you understand why it's there, it becomes quite natural
<Kaufmann> rares, why?
<rares> i look at it as:
<rares> 1. You have a program
<rares> 2. You have functionailyty you want to add
<rares> 3.  You have a resource:
<rares> All the damned libs you installed on your system
<rares> all you need is:
<rares> a language or a natural language parser
<Kaufmann> ?
<rares> that gives you a set amount of flexibility
<Kaufmann> I don't understand what you mean
<rares> the parser interprets your request
<rares> then it searches for all the libraries that give the functionality
<rares> then it calls a few x or gtk calls live to stick stuff up on the screenm ready to go
<Kaufmann> So? What's the problem?
<Kaufmann> Aaaaaaaaaaaaaaah, you mean a problem as in 'issue', not as in 'difficulty'...... yes, of course, I know how to implement it
<rares> problem = exercise
06:30pm
<rares> it's really a matter of getting an elegant system once you've done the basics
<rares> i know it fascinates the fuck out of me
<Fare> hum
<Kaufmann> Well, my design it's pretty elegant... it's based on just a few simple classes who all call do() a lot, and a bit of simple pTk code.
<rares> I figured out one thing though
<rares> eihrul this is right up your alley
<Kaufmann> What?
<rares> separate the form from the function
<rares> that is have generic concepts and specialized interpretations for them
<rares> like a line in a drawing in one context is a path
<rares> in another context it is a channel
<eihrul> Fare: is it worth using indirection cells for GC?
<eihrul> Fare: + copy collector
<rares> have a core that can deal with generic things
<rares> it's CrAZee
<Fare> eihrul: yes, although some do it at page grain
<eihrul> Fare: wouldn't that leek things like mad?
<Fare> also, people tend to favor write-barriers without read barriers
<eihrul> er leak...
<Kaufmann> Behold! We are blessed with the presence of the Faréy One!
<Kaufmann> :)
<eihrul> Fare: yes.. i'm definitely going write barrier...
<Fare> eihrul: if you're implementing GC, you should 1) read the FAQ, 2) subscribe to the gclist
<Fare> K: gakuk!
<eihrul> Fare: i've read most of introductory material on GC at this point...
<eihrul> i just haven't read much on how they perform relative to eachother
<eihrul> Fare: where can i subscribe to the list?
06:40pm
<eihrul> ?
<Fare> eihrul: see lists.tunes.org/cgi-bin/wilma/gclist
<eihrul> thank you
<Fare> click on the info button at bottom
<Fare> performance is *very* application dependent
<Fare> hardware dependent, also
<Fare> and of course GC-dependent
<Fare> and language-dependent
<Fare> so YMMV
<eihrul> well, i'd settle for a comparison of averages of best-case/worst-case
<eihrul> although best-case is irrelevent almost :)
<eihrul> which would be no gc...
<Fare> the ones with best (known) worst-case behaviors are also the slowest
<Fare> they are the ones with read barriers.
<rares> dude the holy grail of the just do it system aint here yet and tunes fits some part of that system
<Fare> oh, and it depends if you consider overall (batch) performance or interactive (realtime) performance
<eihrul> Fare: well, in system i'm using it, objects should tend to be very-very short lived or very-very long lived
<eihrul> so i'm looking at some type of generational algorithm...
<Fare> in the one case, you're considering mostly gctime/runtime, and in the other, max gc pause
<eihrul> Fare: i definitely need somewhat soft real-time performance
<Fare> eihrul: if you want some high examples, try the OCAML GC and the Boehm GC
<Fare> each in its category is very well optimized
<eihrul> i looked at some of pulsar's papers on ocaml lite gc...
<eihrul> it just looked like a generational copy-collector
<Fare> you mean planar?
<eihrul> er, yes :)
<Fare> don't you misname gods, they mightn't like it...
* eihrul/#tunes bites his tongue.
06:50pm
<eihrul> Fare: oh, old generation is mark-and-sweep... only new generation is copy-collect
<rares> word to the wise
<rares> never never
<Kaufmann> So AOL officially broke into the Brazilian market last Tuesday... already I got two CDs in the mail
<rares> move a locked up netscape window when gnmoe has or enlightnment has translucent windows enabled
<Fare> eih: never having implemented any true GC (only toy), I can't say anything about performance but what I read
<eihrul> Fare: fair enough
<eihrul> Fare: you've still read a good deal more than me though :)
<rares> how do i tell linux to validate the contents of the swap file?
<rares> meaning throw out what ever netscape used to own
<Fare> rares: kill netscape
07:00pm
<rares> netscape was kill 9d already and my swap is still 63M / 64 M full
<rares> I would use mozilla but it acts funny sometimes
<rares> though I suppose mr funny is netscape here
<Fare> rares: you can swapoff ; swapon
<rares> will linux pull aback into RAM what it does need?
<Fare> mozilla, like netscape, is a bloated hellspawn
<Fare> rares: sure, if you have enough RAM, that is
<rares> 13M free
<eihrul> emacs syndrome...
<rares> well you know why
<Fare> nmbd put my server to its knees
<rares> emacs syndrome exists?
<Fare> emacs gets you a few tens of megs, once per user (but I'm the only user, so that's fine)
<Kaufmann> rares, because people like it?
<rares> becuase 90% of the popuilation has no clue what software is
<rares> they don't know what's on a system until they see a window
<Fare> rares: 90% of the population has no clue.
<hcf> s/90%/99%/
<rares> hence the C-ROM tray/Coffe-cup holder thingy
<rares> CD even
<eihrul> Fare: what would be the best way to represent strings to take advantage of GC?
<Fare> hcf: ok, then let's go all the way and say: "no one but me has any clue"
<Fare> eihrul: Boehm had a "cords" library as a demo to his GC
<Fare> eihrul: again, it all depends on your application:
<Fare> what will you be doing with those damn strings?
<Kaufmann> No one but me has any clue!
* Kaufmann/#tunes is leaving now
<Kaufmann> buh-bye
<Fare> K: kukag!
<Kaufmann> right back atcha, Fare :)
<rares> .c ya
<rares> anyhow
-:- SignOff Kaufmann: #TUNES (A stranger is just an asshole you haven't met yet.)
<rares> 1% of 6B is 60Million get more than number of Linux users and i'll be amazed
<rares> at the state of  the world's cumulative or average intelligence
* Fare/#Tunes cp -a 2GB worth of debian install over NFS...
<rares> how long did it take?
<Fare> rares: I'll tell you in a few hours
<rares> heh
<Fare> rares: intelligence doesn't accumulate
<eihrul> Fare: well, its a mud server...
07:10pm
<rares> i know it's discovered
<eihrul> Fare: so strings get flung around fairly often
<rares> hehe
<Fare> eihrul: then you might as well use the Boehm GC (or use OCAML) and focus on muddy stuff...
<eihrul> there's no fun in that....
<Fare> eihrul: no, I mean, how do you combine strings? do you often copy them? take substrings? drop them? etc, etc.
<Fare> or is it just "printf and drop" kind of combination?
<rares> all og the above on a mud
<eihrul> well, strings will be concatenated, exploded, and sliced a lot
<rares> you know I should build a mud coprocessor handles strings like there's no tomorrow
<Fare> one way might be to not manipulate the actual strings much, but rather trees of pointers to (sub)strings...
<rares> the problem is
<eihrul> Fare: that was what i was thinking, linked list of chunks of the string...
<rares> people are free to type what ever they want
<Fare> when printed, a string expands to its flattened contents
<Fare> do you often compare strings?
<Fare> that's the big one
<eihrul> not quite as often as the other three operations, but yes
<Fare> also, do you often modify existing strings? insert/delete characters?
<rares> users do that
<Fare> do you compare big strings?
<rares> fat fingers
<eihrul> Fare: strings will be concatenated more than they will be directly modified
<rares> quick tempers
<eihrul> Fare: and the comared strings will typically be small (a single English word)
<Fare> rares: is this editing limited to irc-like input only, or does it affect existing (and possibly shared) strings?
<Fare> eihrul: then you might be better off with trees of symbols as string representation...
<rares> well so far there's no mud where you can paint a house with graffity and have its attributesd be changed to reflect that clearly and precisely
<eihrul> Fare: so when say, exploding a particular string, replace it with a node and its component parts?
-:- SignOff mibin: #TUNES (Ping timeout for mibin[62.10.7.155])
07:20pm
-:- SignOff rares: #TUNES (Read error to rares[wtrb-sh4-port189.snet.net]: EOF from client)
<Fare> eihrul: not sure what you mean
07:30pm
<Fare> rares: if all strings, once entered, are then read-only, things are quite simple
<Fare> rares: approx speed is 20MB/min (that's 330kB/s)
<Fare> so the 2G will be less than 2 hours
* Fare/#Tunes goto 3053
07:40pm
-:- rares [rares@wtrb-sh4-port170.snet.net] has joined #tunes
<rares> opkay whose the bloke who said sawpon/swapoff :)
<eihrul> not me...
07:50pm
* eihrul/#tunes ponders what evil deity cursed us with sleep.
<rares> melatonin
<rares> serotonin
<rares> now i'd like some endorphin
-:- SignOff hcf: #TUNES (Leaving)
<eihrul> agg, kill them!
08:00pm
<rares> moz is not ready yet
08:10pm
-:- SignOff rares: #TUNES ([x]chat)
-:- air [brand@p0wer.qzx.com] has joined #tunes
-:- rares [rares@wtrb-sh2-port184.snet.net] has joined #tunes
<rares> monsters come back to haunt you
<rares> sheesh
<eihrul> yo
<rares> check the F-CPU windows
09:30pm
-:- hcf [nef@me-portland-us900.javanet.com] has joined #tunes
-:- liqf451 [matt@a204b210n109client230.hawaii.rr.com] has joined #tunes
-:- SignOff hcf: #TUNES (Leaving)
-:- liqf451 [matt@a204b210n109client230.hawaii.rr.com] has left #tunes []
[msg(TUNES)] newlog 1999.1121
IRC log ended Sun Nov 21 00:00:01 1999